MISSILE AUTOPILOT DESIGN USING A NEW LINEAR TIME-VARYING CONTROL TECHNIQUE

Design and simulation case studies of a missile autopilot for angle of attack and normal acceleration tracking using a recently developed ex tended-mean assignment (EMA) central technique for linear time-varying (LTV) systems are presented. The EMA control technique is based on a new eigenvalue concept, called series D (SD) eigenvalue, for LTV syste ms. Closed-loop stability is achieved by the assignment of the extende d-mean of these time-varying SD eigenvalues to the left-half complex p lane in a way similar to the eigenvalue (pole) assignment technique fo r linear time-invariant systems. Salient features of the tracking cont roller include the following: 1) good tracking performance for arbitra ry trajectories without scheduling of any constant design parameters t hroughout the entire Mach operating range; 2) time-varying EMA command , or pole locations, to improve tracking performance; 3) implementatio n of the inverse pitch dynamics using a static neural network; and 4) a time-varying bandwidth command shaping filter that effectively reduc es the actuator rate while maintaining good tracking response for both smooth and abrupt guidance command trajectories. Although the autopil ot was designed only for nominal aerodynamic coefficients, excellent p erformance was verified by simulation for +/- 50% variations in the ae rodynamic coefficients.